A dose-escalation study was performed in 27 patients to determine the highest safe and potentially optimal dose of naloxone for treatment of acute cerebral ischemia. All patients received a bolus of naloxone followed by a continuous 24 hour infusion at an hourly rate 50% of the bolus. Loading doses ranged from 2.5 to 200 mg/m2. Total patient doses ranged from 52.3 to 4978 mg. No major dose-related side effects occurred. This study was not designed to determine naloxone's efficacy in stroke, but transient or sustained improvement was observed in 13 patients. Three patients' neurological condition worsened within three hours after the naloxone was discontinued. Our experience suggests that further therapeutic trials of naloxone are worthwhile.